Cloudways MCP Server

A Model Context Protocol (MCP) server for managing Cloudways infrastructure. Provides a secure, modular interface to the Cloudways API with 43+ tools covering server and application management.

Features

• Cloudways API Coverage: 43+ tools organized into logical categories

• Modular Architecture: Separation of concerns with dedicated modules for different functionalities

• Security: Multi-layered authentication with credential encryption, session isolation, rate limiting, and audit logging

• Performance: Asynchronous design with Redis caching and HTTP connection pooling

• Production Ready: Structured logging, error handling, and monitoring

• MCP Compatible: Full Model Context Protocol support for AI assistant integration

Tools Overview

Basic Operations (18 tools)

• Authentication & Info: ping, customer_info, rate_limit_status

• Server Management: list_servers, get_server_details

• Application Management: get_app_details, get_app_credentials, get_app_settings, get_app_monitoring_summary

• Project Management: list_projects, list_team_members, get_alerts

• Infrastructure Discovery: get_ssh_keys, get_available_providers, get_available_regions, get_available_server_sizes, get_available_apps, get_available_packages

Server Operations (12 tools)

• Power Management: start_server, stop_server, restart_server

• Backup & Recovery: backup_server, get_server_settings

• Storage Management: get_server_disk_usage, optimize_server_disk

• Service Control: get_server_services_status, change_service_state

• Caching: manage_server_varnish

• Monitoring: get_server_monitoring_detail, get_server_analytics

Application Management (8 tools)

• Deployment: clone_app, backup_app, restore_app, rollback_app_restore

• Performance: clear_app_cache, get_app_varnish_settings, manage_app_varnish

• Configuration: reset_app_file_permissions, enforce_app_https

• Domain Management: update_app_cname, delete_app_cname

• Analytics: get_app_analytics_traffic, get_app_analytics_php, get_app_analytics_mysql

Security & Access Control (5 tools)

• IP Management: get_whitelisted_ips_ssh, get_whitelisted_ips_mysql, update_whitelisted_ips

• Security Monitoring: check_ip_blacklisted

• Tool Access: allow_ip_siab, allow_ip_adminer

• SSL Management: install_ssl_certificate, remove_ssl_certificate, install_letsencrypt, renew_letsencrypt, set_letsencrypt_auto_renewal, revoke_letsencrypt

• Git Deployment: generate_git_ssh_key, get_git_ssh_key, git_clone, git_pull, get_git_deployment_history, get_git_branch_names

Architecture

Installation & Setup

Prerequisites

• Python 3.11+

• Redis server (for caching and session management)

• Cloudways API credentials

1. Configure environment variables:

   # Generate encryption key for secure credential storage export ENCRYPTION_KEY=$(python -c 'from cryptography.fernet import Fernet; print(Fernet.generate_key().decode())') # Redis configuration export REDIS_URL="redis://localhost:6379/0" # Optional performance tuning export REDIS_POOL_SIZE="500" export HTTP_POOL_SIZE="500" export RATE_LIMIT_REQUESTS="90" export RATE_LIMIT_WINDOW="60"

Encryption Key Management

The server requires a Fernet encryption key for secure credential storage.

Generating a New Key

Exporting an Existing Key

Key Validation

Important Notes

• Key must be 44 characters long (base64-encoded)

• Generated using Fernet.generate_key()

• Do not commit keys to version control

• Changing the key invalidates existing encrypted data in Redis

3. Start the server:

   python main.py

   Server will be available at: http://localhost:7000/mcp

Authentication

The server uses header-based authentication with encrypted credential storage:

Security Features

• Credential Encryption: API keys encrypted with Fernet before Redis storage

• Session Isolation: Unique customer ID generation prevents session cross-contamination

• Token Auto-Renewal: OAuth token refresh before expiration

• Rate Limiting: Token bucket algorithm (90 requests/60 seconds by default)

• Input Validation: Parameter validation with range checks

• Audit Logging: Request/response logging with structured data

Configuration Options

Monitoring & Observability

Structured Logging

• Performance Metrics: Request timing and resource usage

• Security Events: Authentication failures, rate limiting

• API Interactions: Request/response logging with sanitized credentials

• Error Tracking: Error context and stack traces

Health Checks

• ping - Basic connectivity and authentication test

• rate_limit_status - Current rate limit status

• customer_info - Session and authentication status

Development & Extension

Adding New Tools

1. Choose appropriate module in tools/ directory

2. Follow existing patterns for error handling and authentication

3. Use type hints with Pydantic models for parameters

4. Include docstrings

5. Test with various authentication scenarios

Tool Development Template